Signal Regulatory Protein α Negatively Regulates β2 Integrin-Mediated Monocyte Adhesion, Transendothelial Migration and Phagocytosis

Background Signal regulate protein α (SIRPα) is involved in many functional aspects of monocytes. Here we investigate the role of SIRPα in regulating β2 integrin-mediated monocyte adhesion, transendothelial migration (TEM) and phagocytosis. Methodology/Principal Findings THP-1 monocytes/macropahges treated with advanced glycation end products (AGEs) resulted in a decrease of SIRPα expression but an increase of β2 integrin cell surface expression and β2 integrin-mediated adhesion to tumor necrosis factor-α (TNFα)–stimulated human microvascular endothelial cell (HMEC-1) monolayers. In contrast, SIRPα overexpression in THP-1 cells showed a significant less monocyte chemotactic protein-1 (MCP-1)–triggered cell surface expression of β2 integrins, in particular CD11b/CD18. SIRPα overexpression reduced β2 integrin-mediated firm adhesion of THP-1 cells to either TNFα–stimulated HMEC-1 monolayers or to immobilized intercellular adhesion molecule-1 (ICAM-1). SIRPα overexpression also reduced MCP-1–initiated migration of THP-1 cells across TNFα–stimulated HMEC-1 monolayers. Furthermore, β2 integrin-mediated THP-1 cell spreading and actin polymerization in response to MCP-1, and phagocytosis of bacteria were both inhibited by SIRPα overexpression. Conclusions/Significance SIRPα negatively regulates β2 integrin-mediated monocyte adhesion, transendothelial migration and phagocytosis, thus may serve as a critical molecule in preventing excessive activation and accumulation of monocytes in the arterial wall during early stage of atherosclerosis.


Introduction
Recruitment of monocytes from circulation to inflamed tissues plays a pivotal role in the initiation and progression of atherosclerosis [1,2,3]. After migrated to lesion region, monocytes are rapidly differentiated into macrophage which engulf lipids and form the fatty streak [4]. Although the mechanisms that govern the delivery of monocytes from circulation to inflammatory site are not fully understood, the process of monocyte diapedesis has been regarded as a multi-step event that is sequentially regulated by a panel of adhesion molecules and signaling pathways. E-and Pselectins are involved in the initial reversible adherence of monocytes to the endothelial cell monolayers [5]. The following firm adhesion is mediated by monocyte b 2 integrins,including CD11a/CD18 and CD11b/CD18,that recognize vascular cell adhesion molecule-1(VCAM-1) and intercellular adhesion molecule-1 (ICAM-1) on endothelial cells [6]. Firm adhesion of monocytes requires activation of integrins, which can be triggered by agonist-induced activation of G protein-coupled chemokine receptors [7]. Monocytes express CC chemokine receptor 2 (CCR2), which binds monocyte chemoattractant protein-1 (MCP-1), leading to b 2 integrin-mediated firm adhesion and subsequent transmigration of adhered monocytes through the vascular endothelium [8].
Recently signal regulatory protein a (SIRPa) (also designate as SHPS-1 [9], p84 [10], BIT [11], MFR [12], MyD-1 [13], etc.) has been reported to serve as an important modulator for controlling leukocyte inflammatory responses [14,15]. As an immunoglobulin superfamily member (IgSF), SIRPa is expressed mainly by myeloid. SIRPa has a long intracellular domain that contains four tyrosine residues to form two immunoreceptor tyrosine-based inhibition motifs (ITIMs) and this type of signaling structure is highly conserved between mice, rats and humans. Studies have suggested that binding of SIRPa with its extracellular ligand CD47 results in phosphorylations of SIRPa ITIMs, which in turn, leads to their association with SH2-domain-containing protein tyrosine phosphotases SHP-1 and SHP-2 [16,17] to delivers signals that regulate a variety of cellular functions [14]. Ligation of SIRPa by antibody or CD47 recombinant inhibits many leukocyte functions, including phagocytosis [18,19], tumour-necrosis factor production [20] and in vitro transmigration [21,22]. Activation of SIRPa by arterial elastic laminae also inhibits monocyte adhesion [23]. Fibroblasts expressing a SIRPa mutant lacking ITIMs-containing cytoplasmic tail showed increased formation of focal adhesions and actin stress fibres in response to interaction with extracellular matrix, suggesting that SIRPa also plays a role in integrinmediated cytoskeletal organization [24]. Negative regulatory role of SIRPa has also been found in tumor metastasis, survival, and cell transformation [25].
In the present study, to further explore the negative regulatory role of SIRPa in various functional aspects of monocytes, we examined the correlation between expression level of SIRPa in THP-1 cells and THP-1 cell transmigratory capacity. By overexpressing SIRPa in THP-1 cells, we also determined the alteration of b 2 integrin expression and b 2 integrins-mediated cellular functions of monocytes in response to chemoattractant stimulation.

Immunofluorescence, Confocal Microscopy and Flow Cytometry
Surface expressions of b 2 -integrins were detected using flow cytometry as described [8,29]. The relative surface expression was estimated by subtracting the mean fluorescence intensity (MFI) of cells labeled with the nonspecific antibody from that of cells labeled with the antibodies detecting b 2 -integrins. All studies consisted of at least three independent experiments. Flow cytometry was performed and data were analyzed using CELLQUEST software (BD Biosciences). The polymerization of actin filaments in THP-1 cells, induced by pretreatment with 10 nM MCP-1 or 25 ng/mL TNFa for 30-60 minutes, was determined using rhodamine-conjugated phalloidin staining (Molecular Probes) according to the manufacturer's protocol. Briefly, cells treated with cytokines were fixed with 3.7% paraformaldehyde in PBS for 5 minutes, gently permeabilized with 0.1% Triton X-100, and blocked with 1% BSA in PBS for 30 min followed by rhodamine-conjugated phalloidin staining. In some experiments, cells were treated with cytochalasin D (Sigma-Aldrich) to inhibit actin polymerization [30]. Coverslips were mounted with antifade mounting medium (Molecular Probes). Images were captured and analyzed by a laser scanning confocal microscope equipped with an image processing system (Olympus Microsystems).

Transendothelial Migration (TEM)
Migration of THP-1 cells across TNFa-pre-activated HMEC-1 monolayers was performed as previously described [30] with minor modification. Prior to migration assay, HMEC-1 monolayers cultured on gelatin (Difco)-coated transwell filters were treated with 25 ng/mL TNFa for 6 h. THP-1 cells (5.0610 5 /per well) were added to the upper chamber of Transwell inserts containing 200 ml HBSS. 700 ml HBSS containing 10 nM MCP-1 was placed in the bottom chamber. After 90 min and 180 min incubation at 37uC under 5% CO 2 , cells that had transmigrated to the lower chamber were harvested in 1 ml PBS containing 0.1% BSA and labeled with phycoerythrin (PE)-conjugated anti-human CD14 antibody. 10 6 FITC-conjugated standard beads (PharMingen, La Jolla, CA) were added to the cell suspension and the number of THP-1 cells was counted until 10,000 beads were counted by flow cytometry. All experiments were repeated as triplicated fashion in at least three independent studies.

Phagocytosis of Fluorescein Conjugated Bacteria
THP-1 cells transfected with SIRPa or Mock vector were incubated for 3 h with 100 ml of fluorescein-conjugated E. coli K-12 bioparticles (Molecular Probes) [33]. The E. coli suspension was aspirated, and after three washes with Hank's balanced salt solution devoid of Ca 2+ and Mg 2+ (HBSS 2 ), cell associated fluorescence was observed under microscopy or measured using a SpectraMax microtiter plate reader (Molecular Devices).

Western Blot
THP-1 cells were solubilized in lysis buffer containing 1% Triton X-100 and a panel of protease inhibitors at 4uC. Pellet was removed after centrifuged at 13,0006g for 5 minutes. Supernatant was normalized for total protein, and loaded on 10% SDS-PAGE. After electrophoresis and transfer onto Hybond membranes, membranes were blocked with 5% non-fat milk. Antigens were detected using suitable primary antibodies followed by incubation with HRP-conjugated antibodies and ECL (Amersham) detection.

Statistical Analysis
Data were analyzed by the Student t test; P values of ,0.05 were regarded as significant differences (*, p,0.05; **, p,0.01).

Downregulation of SIRPa in THP-1 cells is correlated with enhanced b 2 integrin-mediated cell adhesion
It has been reported that advanced glycation end products (AGEs) are involved in tissue damage associated with diabetic complications and aging [34,35]. Although the mechanism is still not clear, monocytes tend to be activated by AGEs and show an enhanced chemotaxis under such inflammatory conditions. As shown by Western blot analysis in Figure 1A, the expression of SIRPa in THP-1 cells was decreased after AGEs treatment. Served as controls, b-actin level was not altered. The downregulation of SIRPa in AGEs-treated THP-1 cells is contrast to that of receptor for advanced glycation end products (RAGE) and junctional adhesion molecule-like protein (JAML), which expression levels are both increased after AGEs treatment (Zen et al, unpublished). Fig. 1B showed the quantitative analysis of SIRPa downregulation by AGEs in a dose-dependent fashion. Interest- ingly, AGEs-treated THP-1 cells showed a significant enhanced cell surface expression of b 2 integrins, in particular CD11b/CD18, in response to the stimulation of MCP-1 (Fig. 1C). Also, compared to BSA-treated THP-1 cells, AGEs-treated THP-1 cells had a higher percentage of cell adhesion to TNFa-activated HMEC-1 monolayer (Fig. 1D). In response to MCP-1, AGEs-treated THP-1 cells also showed an increased transmigration across TNFaactivated HMEC-1 monolayers compared to THP-1 cells treated with BSA (Fig. 1E). Together, these results suggest that AGEs treatment can activate THP-1 cells and enhance cell chemotaxis.

Overexpression of SIRPa in THP-1 cells inhibits MCP-1induced cell surface expression of b2 integrins
To define the role of SIRPa in regulating monocyte inflammatory response, we characterized the alteration of b 2 integrin expression and b 2 integrin-mediated cell adhesion, migration and phagocytosis in THP-1 cells after significantly increase SIRPa expression level. As shown in Figure 2A, immunoblot analysis showed that the delivery of the pcDNA3.1 vector encoding human SIRPa into THP-1 cells profoundly enhanced SIRPa expression. Compared to mock-transfected THP-1 cells, SIRPa-transfected THP-1 cells also showed a significantly enhanced expression of SIRPa on cell surface, as indicated by flow cytometry (Fig. 2B).
Similar to circulating monocytes, THP-1 cells normally express b 2 integrins [8,36,37] and their cell surface expression levels are rapidly up-regulated by chemoattractants during inflammatory response. We next determined the alteration of cell surface expression of CD11b/CD18 and CD11a/CD18 in SIRPa-or mock-transfected THP-1 cells. In these experiments, SIRPa-or mock-transfected THP-1 cells were treated with or without 10 nM MCP-1 for 30 min and then directly labeled with PE-conjugated mouse IgG specific for human CD11a, CD11b, and CC chemokine receptor 2 (CCR2), and surface expression was analyzed using flow cytometry. The results showed that SIRPa overexpression in THP-1 cells did not affect the basal level of b 2 integrin expression on cell surface but significantly reduced the up-regulation of cell surface b 2 integrin expression by MCP-1 (Fig. 2C).

SIRPa overexpression affects CD11b/CD18-mediated THP-1 cell functions in response to MCP-1 stimulation
Chemokines such as MCP-1 has been reported to trigger integrin-mediated firm adhesion and subsequent transmigration of monocytes [8]. As shown in Figure 3A, MCP-1-stimulated THP-1 cells showed a significant integrin-mediated firm adhesion to TNFa-activated HMEC-1 monolayers. However, this firm adhesion was largely reduced in THP-1 cells with SIRPa overexpression. Since TNFa-stimulated HMEC-1 monolayers express both VCAM-1 and ICAM-1, the specific ligands for b 1 and b 2 integrins, respectively, additional adhesion assays were performed using plates coated with human recombinant ICAM-1 or VCAM-1, respectively. To estimate background adhesion, control adhesion assays were performed using THP-1 cells pretreated with Bt2cAMP, a permeable analogue of cAMP that blocks integrin-dependent firm adhesion triggered by MCP-1 [30]. SIRPa-transfected THP-1 cells did not show MCP-1-induced firm adhesion to plates coated with ICAM-1, while adhesion to plates coated with VCAM-1 was intact (Fig. 3B).
Previous studies have reported that chemokine-mediated activation of b 2 integrins was essential for THP-1 cell adhesion and subsequent transmigration through endothelial monolayers [7,8,38]. Therefore, the effect of SIRPa overexpression on transendothelial migration (TEM) of THP-1 cells was examined by transmigration assay. In mock-transfected THP-1 cells, MCP-1 triggered strong THP-1 cell migration across HMEC-1 monolayers. As shown in Figure 4, more than 20% of total applied THP-1 cells were migrated across TNFa-activated HMEC-1 monolayers after 3 h incubation. In contrast, MCP-1-triggered transmigration of THP-1 cells that were overexpressed with SIRPa was strongly reduced (Fig. 4). In both mock-transfected and SIRPa overexpressed THP-1 cells, spontaneous migration of THP-1 cells in the absence of MCP-1 was minimal (data not shown).
Leukocyte b 2 integrin-mediated cell firm adhesion initiates cell shape changes and spreading of monocytes, events that must occur for subsequent cellular locomotion and transmigration. Next we investigated the effect of SIRPa overexpression on TNFa-and MCP-1-stimulated actin polymerization and cell spreading in THP-1 cells. SIRPa-or mock-transfected THP-1 cells were stimulated with TNFa or MCP-1 for 30 minutes, then fixed, and labeled with rhodamine-conjugated phalloidin to visualize actin filaments. As a control, mock-transfected THP-1 cells were pretreated with cytochalasin D to inhibit actin polymerization [39]. Labeled monocytes were mounted on coverslips and images were obtained using confocal microscopy. As shown in Figure 5, confocal microscope images showed that mock-transfected THP-1 cells exposed to TNFa or MCP-1 underwent morphological changes resulting in multiple pseudopods (arrowheads) with abundant actin filaments, and that this process was inhibited by cytochalasin D. In contrast, significantly less TNFa-or MCP-1stimulated actin polymerization and cell spreading had occurred in THP-1 cells with SIRPa overexpression (Fig. 5).
The phagocytic function of THP-1 cells is also dependent on b 2 integrins [40,41,42]. Next we examined the effect of SIRPa overexpression on the capacity of THP-1 cells to engulf fluorescein-labeled E. coli K12 bioparticles. As shown by confocal images in Figure 6A, the mock-transfected THP-1 cells showed a   significant phagocytosis of fluorescein-conjugated bacteria particles after 3 h incubation (Fig. 6A, arrows), while in SIRPatransfected THP-1 cells, uptake of fluorescein-conjugated bacteria particles was strongly reduced. The quantitative analysis of uptaking fluorescein-conjugated bacteria particles by THP-1 cells was shown in Figure 6B. Taken together, these results clearly show that SIRPa overexpression in THP-1 cells reduces various inflammatory responses mediated by leukocyte b 2 integrins.

Discussion
Recent studies have demonstrated that SIRPa is involved in regulating various inflammatory responses of leukocytes, in particular leukocyte chemotaxis and phagocytosis. By studying the leukocyte b 2 integrin-mediated functional changes in THP-1 cells after downregulation or overexpression of SIRPa level, we show that SIRPa negatively regulates b 2 integrin-mediated THP-1 cell inflammatory responses, such as adhesion, transendothelial migration and phagocytosis.
Correlation between SIRPa protein level and CD11b/ CD18-mediated cellular functions in THP-1 cells Ligation of SIRPa with its extracellular ligand CD47 results in phosphorylations of SIRPa ITIMs, which in turn, leads to their association with SH2-domain-containing protein tyrosine phos-photases SHP-1 and SHP-2 [16,17] to delivers signals that regulate a variety of cellular functions [14]. Binding of SIRPa by antibody or CD47 recombinant inhibits many leukocyte functions, including phagocytosis [18,19], tumour-necrosis factor production [20] and in vitro transmigration [21,22]. Activation of SIRPa by arterial elastic laminae also inhibits monocyte adhesion [23]. Fibroblasts expressing a SIRPa mutant lacking ITIMs-containing cytoplasmic tail showed increased formation of focal adhesions and actin stress fibres in response to interaction with extracellular matrix, suggesting that SIRPa is also involved in mediating outside-in signal transduction during cell-matrix interaction. Using THP-1 cell as model cell line, here we show that SIRPa protein level is downregulated by AGEs treatment, which is also correlated to an enhanced cell surface expression of b 2 integrins and b 2 integrins-mediated cell adhesion (Fig. 1). The finding of SIRPa reduction in AGEs-treated THP-1 cells is supported by a recent report that mouse macrophages have lower SIRPa expression level following LPS stimulation [43]. The correlation between SIRPa expression level and chemoattractant-induced cell surface upregulation of b 2 integrins and b 2 integrins-mediated THP-1 cell inflammatory responses is further characterized in THP-1 cells overexpressed with SIRPa ( Fig. 2-6). The results not only confirm the inhibitory function of SIRPa on THP-1 inflammatory responses, but also indicated that the role of SIRPa in THP-1 cells is through affecting the functions of b 2 integrins, particularly CD11b/CD18. It is worthy to note that overexpression of SIRPa does not alter the basal level of b 2 integrin expression but the upregulation of b 2 integrins by MCP-1 stimulation, suggesting that SIRPa is one of essential molecules along the signal pathways that may regulate the synthesis, transportation and translocation process of b 2 integrins. Moreover, if AGEs and other inflammatory factors can affect b 2 integrin expression and function through down-regulating SIRPa, it might be reasonable to conclude that SIRPa can mediate an insideout signal in regulating b 2 integrin function.

SIRPa as a negative regulator in monocyte recruitment during inflammation
The expression of b 2 integrins and adhesion molecules in monocytes is regulated by chemokines such as MCP-1, SDF-1 alpha and RANTES [32,44,45,46]. The positive correlation between CD11b expression in circulating monocytes and the degree of monocyte infiltration into the proatherogenic vascular wall has been well-documented [8,47,48]. The increased expression of monocyte CD11b under pro-inflammatory conditions enhanced MCP-1-mediated chemotaxis in vitro [8], induced excess monocyte adhesion to vascular endothelium, and increased formation of neointima and atherosclerotic plaques [48]. Although SIRPa overexpression did not affect surface expression of CCR2, the receptor for MCP-1, it resulted in a profound reduction of MCP-1-mediated upregulation of THP-1 cell cell surface b 2 integrins and THP-1 cell TEM. In addition to reduction of CD11b and other b 2 integrins, our study has also demonstrated that overexpressing SIRPa in THP-1 cells display less cell spreading and actin polymerization in response to chemokine stimulation. The mechanism by which SIRPa modulates chemokine-induced cell spreading and actin polymerization is unknown although several possibilities exist: a) directly activates protein phosphatase and initiates signal pathways that attenuate filament actin polymerization and cell spreading, and b) binding to integrinassociated protein CD47 and modulating the integrin functions. Since SIRPa is a cellular ligand of CD47, which can augment the functions of integrins of the b 1 , b 2 and b 3 families via initiating heterotrimeric Gi protein signaling [49], thus modulating a range of cell activities including cell motility and adhesion, and leukocyte adhesion, migration and phagocytosis. Indeed, phagocytosis of bacteria by THP-1 cells, an event that is largely dependent on b 2 integrin and actin polymerization, was significantly reduced by overexpression of SIRPa. This result was in agreement with the previous finding that SIRPa contributes to down-regulating the macrophage phagocytic response [18]. In summary, the present study demonstrates for the first time that SIRPa overexpression potently inhibits the various inflammatory responses of THP-1 monocytes/macrophages mediated by b 2 integrins. The induction of SIRPa expression in THP-1 cells led to a reduction of chemokine-induced cell surface expression of b 2 integrins, which eventually resulted in less cell adhesion, cellular spreading, cell transmigration and phagocytosis. This observation suggests that SIRPa may function to decrease transendothelial migration of monocytes or other circulating leukocytes, reduce the burden of inflammatory cells in atheroma, and ultimately decrease plaque mass under atherogenic conditions. Since migration of monocytes across blood vessel lining endothelial monolayers is a key component during early stage of atherosclerosis, such an outcome would indicate that SIRPa overexpression in monocytes or macrophages has an anti-atherogenic effect and that SIRPa is a potential target in therapeutical implications.